unsubscribe

2006-05-12 Thread Rowan O'Donoghue
kind regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Unit 2,
First Floor,
Bracken Court,
Sandyford Industrial Estate,
Dublin 18.

DDI: + 353 - 1 - 2999357
Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
http://www.unitech-ie.com

Unitech Systems announces partnership with Commvault Systems:
http://www.unitech-ie.com/view_pr_article.php?pr_id=31


Question on HSM TSM.

2004-10-19 Thread Rowan O'Donoghue
All,

In an environment where you are using HSM to manage disks (say for example
Windows drives) after a period of time the file will be migrated to a
storage pool within TSM and a stub file is left on the filesystem as a
marker.

I'm assuming that when this migration happens the TSM backup/archive client
identifies the change and only backs up the stub - or does it need to
recall the file from HSM and then back it up? If it does only backup the
stub the issue is that you do not have a copy of a real file only a stub
and no data.

Can someone give me an idea as to how they have this type of scenario
implemented within their environment?


kind regards,

Rowan.


Re: Question on HSM TSM.

2004-10-19 Thread Rowan O'Donoghue
Richard,

I agree with what you say, and it is pointless in restoring the file to the
original filesystem just to back up that file if it is already migrated to
tape. I see that there is an option with DiskXtender to do a backup of the
file to TSM before it is migrated to tape and any subsequent backups of the
file can be ignored if the file size is zero

Still reading ;)

kind regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Unit 2,
First Floor,
Bracken Court,
Sandyford Industrial Estate,
Dublin 18.

DDI: + 353 - 1 - 2999357
Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com
|-+---|
|   Richard Sims [EMAIL PROTECTED] |   |
|   Sent by: ADSM: Dist Stor |   |
|   Manager  | To|
|   [EMAIL PROTECTED]|   [EMAIL PROTECTED]|
| |   M.MARIST|
|   19/10/2004 12:17  |   .EDU|
| | cc|
| |   |
| Please respond to   |Subject|
| ADSM: Dist Stor|   Re: |
| Manager|   Question|
|  [EMAIL PROTECTED] |   on HSM |
| |   TSM.|
| |   |
| |   |
| |   |
| |   |
| |   |
| |   |
|-+---|






On Oct 19, 2004, at 6:58 AM, Rowan O'Donoghue wrote:

 In an environment where you are using HSM to manage disks (say for
 example
 Windows drives) after a period of time the file will be migrated
 to a
 storage pool within TSM and a stub file is left on the filesystem as a
 marker.

 I'm assuming that when this migration happens the TSM backup/archive
 client
 identifies the change and only backs up the stub - or does it need to
 recall the file from HSM and then back it up? If it does only backup
 the
 stub the issue is that you do not have a copy of a real file only a
 stub
 and no data.

 Can someone give me an idea as to how they have this type of scenario
 implemented within their environment?

Rowan - TSM never backs up HSM stub files, as that would be pointless.
               It backs up the file, either from the file system or
directly
from the HSM storage pool if migrated and using same server.

See the (old) HSM redbook at
 http://www.redbooks.ibm.com/abstracts/sg244631.html
which explains how it all works in different configurations, and advises
on restoral methods.

  Richard Sims    http://people.bu.edu/rbs

ForwardSourceID:NT00015B9E

IBM 3584 Problems (tapes getting stuck in control path drive) with Solaris

2003-11-10 Thread Rowan O'Donoghue
Hi,
 
I noticed a while back that a few people were having problems with the IBM3584's and 
some flavours of Unix. The problems mainly surrounded the issue of LTO tapes getting 
stuck in the drive that is also the control path for the library.
 
I'm having the same problem now with both of my Solaris servers and my two 3584 
libraries. I have G2 drives at 2CKH firmware and when I perform an audit libr the tape 
that gets placed into drive 1 (element 257) always gets stuck and never ejects. I've 
also had scenarios where I have had to get an engineer in to remove the tape manually. 
I am going via SAN and am sharing the tape drives out to Solaris and Windows systems. 
Everything seems to be going to plan so far (apart from the performance on the Solaris 
SAN clients) and LAN-free backups are working perfectly. I have noticed the tapes to 
get stuck in the same drives when performing a backup stg from one library to another.
 
Anyone ever know what resolved this? I'm beginning to tear my hair out! I've just 
pulled down the latest IBMtape device drivers for Solaris (v4.0.8.0) but still no joy.
 
Anyone?


regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel: + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com


Shared mem with Solaris LAN-Free

2003-11-06 Thread Rowan O'Donoghue
All,
 
I keep getting the error when trying to run the sharedmem commethod on the TSM Storage 
Agent on my Sun Solaris server.
 
ANR9600E Failed to allocate memory for shared memory communications. Error: EINVAL.
ANR8289W Shared Memory driver is terminating due to error in creating a new thread.

Anyone got any ideas on how to resolve this? The server has got 8GB of RAM!

regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel: + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com


TSM 5.2 Sun Solaris - Language Format problems.!??!!!

2003-10-29 Thread Rowan O'Donoghue
Had TSM 5.1 installed with no problems in this environment before.
 
Using Solaris 8, and TSM 5.2
 
Upon installing the TSM 5.2 server code I get the following errors:
 
Error opening catalog /opt/tivoli/tsm/server/bin/dsm/en_IE.ISO8859-15/en_IE.ISO8
859-15/en_IE.ISO8859-15/en_IE.ISO8859-15/en_IE.ISO8859-15/C.cat, for language /e
n_IE.ISO8859-15/en_IE.ISO8859-15/en_IE.ISO8859-15/en_IE.ISO8859-15/en_IE.ISO8859
-15/C
ANR0915E Unable to open language /EN_IE.ISO8859-15/EN_IE.ISO8859-15/EN_IE.ISO8
8859-15/EN_IE.ISO8859-15/EN_IE.ISO8859-15/C for message formatting.
 
I assumed that these would just be warning errors, but after using DSMFMT to format my 
database and log volumes I cannot use the DSMSERV FORMAT commant as it comes back with 
a message saying that the format parameters are missing!!!??!!!
 
Anyone else having this problem?
 
I'm using: DSMSERV FORMAT 1 /opt/tivoli/tsm/server/bin/log1.dsm 1 
/opt.tivoli/tsm/server/bin/db1.dsm 
 
The DSMFMT command was tried first, but no joy.
 
Is the error in relation to the language the one causing the problems? if so, I'm 
gonna have a problem as the locale on all these servers has to stay EN_IE (Irish).
 
Anyone? Maybe one for the SUN gurus!
 
R.
 
 


Two TSM Servers with a Storage Agent?

2003-10-23 Thread Rowan O'Donoghue
Hi guys,
 
I want to be able to configure a server using the LAN-free storage agent to be able to 
back up to storage pools on two different TSM servers.
 
I.e. It will backup direct to tape via the SAN to a backup storage pool defined on one 
TSM server, and also backup direct to tape via the SAN to an ARCHIVE storage pool 
define on another TSM server.
 
Can I just define two tsm storageservers on the storage agent??? Anyone know if this 
is possible?


regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel: + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com


Subfile backups (or Adaptive Copy)

2003-10-14 Thread Rowan O'Donoghue
All,

Have a customer looking at subfile backups for two webservers which are
hosted at a remote hosting facility and connected via a WAN.

We are looking at using subfile backups to enable these servers to
participate within the TSM hierarchy, but I want to get a feel of
real-world-experiences from people who are actually using it in production
at the moment.

What have been your main issues? (apart from the obvious being expanded
TSM processing time on the client, and slow restores).

Also, how have you sized the subfile cache area? Was there a rule of thumb
for this sizing?


thanks!

Rowan.


Re: AW: Subfile backups (or Adaptive Copy)

2003-10-14 Thread Rowan O'Donoghue
Interesting to note about the transfer of the full file after the 32 
subfile backups - does this definitely happen?

You mention that 1GB was used as the subfile cache area...how big was the 
amount of space available on the server?

I believe that the TSM client does not perform housekeeping on this 
directory (i.e. it does not clean out old binary images) and you have to 
do this yourself.

thanks,

Rowan.



Salak Juraj [EMAIL PROTECTED] 
Sent by: ADSM: Dist Stor Manager [EMAIL PROTECTED]
14/10/2003 16:55

Please respond to
ADSM: Dist Stor Manager [EMAIL PROTECTED]


To
[EMAIL PROTECTED]
cc

Subject
AW: Subfile backups (or Adaptive Copy)






I tested it on couple of file servers, it worked like  a breeze,
but at the last I was definitely limited by it´s limits 
(grr, my wording in not worth to loose a word about)
namely subfile cache area - which is 1 GB as far as I can remember,
anyway, far too small to keep all delta files dictated by my file servers 
in
question.
I ended in purchasing extra bandwidth and performing regular forever
incremental.

I still have got a very small client location where subfile backup 
simply works and works - no troubles at all so far.

Another point to watch is that regulary, 
I believe after each 32 subsequent subfile backups, 
full file size will be transfered once again.

regards
juraj salak



-Ursprüngliche Nachricht-
Von: Rowan O'Donoghue [mailto:[EMAIL PROTECTED]
Gesendet: Dienstag, 14. Oktober 2003 17:31
An: [EMAIL PROTECTED]
Betreff: Subfile backups (or Adaptive Copy)


All,

Have a customer looking at subfile backups for two webservers which are
hosted at a remote hosting facility and connected via a WAN.

We are looking at using subfile backups to enable these servers to
participate within the TSM hierarchy, but I want to get a feel of
real-world-experiences from people who are actually using it in production
at the moment.

What have been your main issues? (apart from the obvious being expanded
TSM processing time on the client, and slow restores).

Also, how have you sized the subfile cache area? Was there a rule of thumb
for this sizing?


thanks!

Rowan.

ForwardSourceID:NT00019F26 


TSM DR

2003-10-01 Thread Rowan O'Donoghue
All,


The one thing I've always found to be a downside to TSM is the ability to
easily identify the EXACT amount of tapes required to rebuild an
environment to it's most current state. I have wrote SQL scripts
previously which performs a query against each TSM client and it's
filespaces for those files marked as ACTIVE and then linked to the tape
which held these files.

select distinct(con.volume_name) from contents con,backups bak where
concat(concat(bak.hl_name,' '),
bak.ll_name)=con.file_name and bak.state='ACTIVE_VERSION' and
con.node_name='STEGANOS' and con.type='Bkup'


The trouble with this query/script is that it took forever for it to run
and would not be much use during  a DR scenario.

Is there a simpler way to do this? Again, rather than recall all offsite
pool tapes, I only want the bare minimum to recover the systems to their
most recent state.


regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com


Re: TSM DR

2003-10-01 Thread Rowan O'Donoghue
Agreed, it is probably the most important shortcoming!

In terms of when a DR is performed for a client, surely the TSM client
performs a similar query with the TSM server to identify the files
required to be restored?

Maybe there's a smooth streamlined script for this purpose that IBM just
dont want us to have ;)

Ah well will keep on looking..

regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com



Richard Sims [EMAIL PROTECTED]
Sent by: ADSM: Dist Stor Manager [EMAIL PROTECTED]
01/10/2003 17:35
Please respond to
ADSM: Dist Stor Manager [EMAIL PROTECTED]


To
[EMAIL PROTECTED]
cc

Subject
Re: TSM  DR






The one thing I've always found to be a downside to TSM is the ability to
easily identify the EXACT amount of tapes required to rebuild an
environment to it's most current state. ...

That has been a Holy Grail quest for us customers.  There remains no
feasible
way to get such a list from the product.  This condition probably stems
from
the product philosophy of being file-oriented rather than tape-oriented,
in
that the customer need not be concerned with the current tapes upon which
the files reside.  Obviously, DR needs conflict with that neatness, and so
we all would like to see the product accommodate such needs.

   Richard Sims, BU


Re: Netware 6 TSM 5.2 Client

2003-09-26 Thread Rowan O'Donoghue
Jim,

first of all thanks for the reply.

One question though, these will be shared cluster volumes so what does the
/cluster=off do in the entry below?


regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com



Jim Kirkman [EMAIL PROTECTED]
Sent by: ADSM: Dist Stor Manager [EMAIL PROTECTED]
25/09/2003 19:47
Please respond to
[EMAIL PROTECTED]


To
[EMAIL PROTECTED]
cc

Subject
Re: Netware 6  TSM 5.2 Client






Here you go. A search on nss volumes on search.adsm.org will turn this
up rather quickly.

Novell's TID 10065605 and patch NW6SMS1A.EXE are needed. In particular,
you need the following entry in the smsrun.bas file: nlmArray =
Array(SMDR, TSA600 /cluster=off, TSAPROXY)in order for TSM to
recognize the clustered volumes.


Rowan O'Donoghue wrote:

 Hi all!

 I have a customer who has just implemented Netware 6.0 and the TSM 5.2
 client. Everything seems to be ok apart from the fact that there are SAN
 volumes which are assigned to the Netware 6 server (filespace called
 DATA01:) which are not recognised by the TSM client. the SYS: volume is
 being picked up no problem. Even if I try and perform an INC DATA01:\ it
 comes back telling me that the filespace is invalid.

 Anyone have the same problem?

 All the Netware 5 servers with the TSM 5.1 client are ok - then again, I
 dont have any SAN attached volumes on these servers.

 regards,

 Rowan O'Donoghue
 Technical Services Manager
 Unitech Systems
 Bracken Business Park
 Bracken Road
 Sandyford Industrial Estate
 Dublin 18.

 Tel:  + 353 - 1 - 2942300
 Fax: + 353 - 1 - 2942319
 www.unitech-ie.com

--
Jim Kirkman
AIS - Systems
UNC-Chapel Hill
966-5884


Netware 6 TSM 5.2 Client

2003-09-25 Thread Rowan O'Donoghue
Hi all!

I have a customer who has just implemented Netware 6.0 and the TSM 5.2
client. Everything seems to be ok apart from the fact that there are SAN
volumes which are assigned to the Netware 6 server (filespace called
DATA01:) which are not recognised by the TSM client. the SYS: volume is
being picked up no problem. Even if I try and perform an INC DATA01:\ it
comes back telling me that the filespace is invalid.

Anyone have the same problem?

All the Netware 5 servers with the TSM 5.1 client are ok - then again, I
dont have any SAN attached volumes on these servers.

regards,

Rowan O'Donoghue
Technical Services Manager
Unitech Systems
Bracken Business Park
Bracken Road
Sandyford Industrial Estate
Dublin 18.

Tel:  + 353 - 1 - 2942300
Fax: + 353 - 1 - 2942319
www.unitech-ie.com